论文部分内容阅读
种群比较试验（随机区组设计）一般有三种类型。一种用于比较纯种性能，如品种测验；另一种用于比较不同杂种，如经纬杂交试验；还有一种是参加比较的既有纯种也有杂种，如完全的和不完全的双列杂交试验。本文提出了一套简单易用的公式，以在设计种群比较试验时确定各参试种群的样本大小。考虑了各参试种群的方差的不同和获取每一个体的观测值所需费用的不同。给出了四个应用例子以解释这些公式的用法。
There are generally three types of population comparison trials (randomized block design). One for comparing purebred traits, such as breed tests; the other for comparing different hybrids, such as longitude and latitude crossbreeding; and one for both comparisons of both existing and hybrid breeds, such as complete and incomplete double traits Hybridization test. In this paper, a simple and easy-to-use formula is proposed to determine the sample size of each tested population when designing a population comparison test. The differences in the variances of the tested populations and the cost of obtaining observations for each individual are considered. Four application examples are given to explain the usage of these formulas.